network-credential-cracking
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface where instructions processed from external data could potentially influence agent behavior or tool execution.
- Ingestion points: The skill processes external, potentially attacker-controlled artifacts such as hashes in
hashes.txt, encrypted files likeid_rsa,f.zip,db.kdbx, and content crawled from URLs viacewl(found inSKILL.md). - Boundary markers: Absent; the instructions assume artifacts match expected cryptographic formats without implementing explicit validation or delimiters.
- Capability inventory: The skill utilizes standard subprocess tools including
hashcat,john,cewl, andcrunchto process the ingested data (found inSKILL.md). - Sanitization: Absent; the skill provides no guidance on sanitizing the content of captured artifacts or crawled web data before passing them to CLI tools.
Audit Metadata